Pharo
All integrations

Pharo + Claude MCPLive

Query your live sales data directly inside Claude.

Pharo runs a native Model Context Protocol server. Connect it to Claude Desktop or ChatGPT to query live campaign results, run deal inspection prompts, and analyse call objections in natural language — without exporting anything.

01

Natural language pipeline queries

Ask “which buyer objections came up most on calls this week?” or “summarise account signals for Nimbus Freight.”

Claudepharo · mcp
Which objections came up most this week?
Pricing (14), then already have a dialer (9). Pricing objections were twice as likely on calls under two minutes.
02

Campaign analysis in place

Have Claude read call transcripts and campaign performance together, and suggest script revisions from what actually happened on the calls.

Timelinetoday
Call · 4m 12s · connected

Asked about ramp time for new SDRs. Wants to see it on their own list.

Stage → Meeting booked

Set from the call disposition.

03

No ETL

Connect over standard MCP endpoints in seconds. Nothing is copied into a warehouse first.

Add serverno ETL
pharo mcp add --client claude
Scoped, read-only by default, revocable per client

Three steps, and none of them is a project

1 Add the MCP server
2 Approve the scopes
3 Ask a question

Common questions

What is the Pharo MCP server?

A Model Context Protocol server over your Pharo campaigns, call transcripts and account context. Connect it to Claude Desktop, ChatGPT or another MCP client and you can query live results in natural language without exporting anything into a warehouse first.

What can it read, and can it write anything?

Campaign results, call transcripts and account context, scoped and read-only by default. Access is granted per client and is revocable per client.

Do I need a data warehouse or an ETL job first?

No. That is the point of it — MCP is a standard endpoint the client connects to directly, so nothing is copied anywhere in order to be asked a question.

The rest of the connector set

The orchestration layer runs on Nango, so most of the long tail is configuration rather than engineering. If the one you need is not here, tell us which and it moves up the queue.

Get started

Ready to eliminate dead dials and marked-up credits?

Seven days, 1,000 credits, no credit card. Or thirty minutes with the two people who would do the work.